摘要
获得胎心率信号的方法有多种,而以超声传感器检测法应用最广。然而如何在嵌入式实时系统中快速、准确地从含有噪声干扰的超声多普勒回波信号中提取胎心率信号成为一个难点。针对该难点,提出了一种基于经验模态分解(EMD)与提升小波变换相结合的去噪方法,再将去噪后的胎心信号采用希尔伯特变换提取信号包络,最后通过自相关运算得出胎心率。实验结果表明:采用这种方法能够有效提取到胎心率信号,使得胎心率计算耗时大大减少,同时,计算的准确性得到提高。
There are many ways to obtain FHR signal,and the most common way is to use ultrasonic sensor.However,it is a challenge to quickly and accurately extract pure fetal heart signal from Doppler fetal heart echo signal which has a lot of noise interference in embedded real time system. Considering the difficulty,a new kind of denoising method based on empirical mode decomposition（ EMD） and lifting wavelet transform is constructed.Then use the Hilbert transform to extract envelope of signal and get fetal heart rate through autocorrelation operation. The experimental results show that the algorithm can effectively extract fetal heart rate signal,greatly reduce heart rate calculation time and improved veracity of calculation.
出处
《传感器与微系统》
CSCD
2016年第4期144-147,共4页
Transducer and Microsystem Technologies
基金
国家自然科学基金资助项目(NSFC-ANR项目(中法))(61411136003)
关键词
胎心率
超声传感器
经验模态分解
提升小波变换
自相关
fetal heart rate（FHR）
ultrasonic sensor
empirical mode decomposition（EMD）
lifting wavelet transform
autocorrelation
